We previously developed a potential Drosophila model of mammalian neurodegenerative disease by inactivating the protein kinase Cdk5. This is the fly homolog of one of the two main proteins responsible for phosphorylating tau into the form found in the neurofibrillary tangles that are characteristic of many forms of human neurodegeneration. We showed that there are many phenotypic similarities between the neurodegenerative syndrome in this endogenous process of flies and that observed in humans and mice, thus validating our Drosophila mutant as a model of the mammalian disease process. More recently, we have found that Cdk5 also controls various aspects of developmental disassembly of axons and dendrites, in particular the stability of the microtubule cytoskeleton. Thus, we find that loss of Cdk5 activity delays the onset of developmental remodeling of axons and dendrites of the fly central brain during metamorphosis, while hyperactivation of Cdk5 accelerates remodeling. Genetic interaction and pharmacological experiments reveal that this is mediated partly, but not completely, through Cdk-dependent destabilization of microtubules in these neurons. Given the many anatomic and molecular similarities between developmental remodeling of neurons and their pathological degeneration in disease, it is likely that this function of Cdk5 is directly relevant to the mechanism of neurodegeneration in p35 mutants. We also initiated systematic, genome-wide investigations into the consequences of Cdk5 inactivation. We have used microarrays to perform gene expression profiling of adult flies having, or lacking, Cdk5 activity. We have focused our efforts on early events in degeneration by isolating RNA from young flies, before the onset of overt cellular or organismal defects in the mutants. Preliminary analysis reveals two classes of gene expression changes in mutants lacking Cdk5 activity. We observe alterations in genes directly important for neural function, including sensory transduction, synaptic transmission and ion homeostasis. We also observe changes in the expression of genes associated with a variety of organismal physiological processes we had previously hypothesized to be associated with Cdk5 function, based on our phenotypic analysis. These include proteostasis, redox state, mitochondrial structure and function and stress-sensitivity. The microarray analysis, therefore, serves as a powerful independent validation of the phenotypic analysis. Moreover, as this unbiased, genome-wide array analysis failed to identify major classes of genes not predicted by the earlier studies, these data suggest that we have identified the major functional pathways downstream of Cdk5 that are responsible for its pathological consequences. Remarkably, we find that many of the processes downstream of Cdk5 are also physiological hallmarks of natural aging, and we are currently investigating this connection.
